Vous êtes sur la page 1sur 58

20 December 2017

Project Tenor II

Summary Report
Report Prepared for

The National Bank of Moldova

Nexus Place, 25 Farringdon Street


London EC4A 4AB
T +44 (0)20 7029.5000

kroll.com
Project Tenor II report

Private & Confidential


Restricted Use Warning

This report was prepared by Kroll at the request of the client to whom it is furnished. The client agrees that
reports and information received from Kroll, including this report, are strictly confidential and are intended
solely for the private and exclusive use of the client to understand how a fraud was perpetrated against the
Three Moldovan Banks (as defined below). Any other use (including for employment purposes, credit
evaluation or insurance underwriting purposes) is strictly prohibited and the client has agreed that no such
use will occur. Any communication, publication, disclosure, dissemination or reproduction of this report or any
portion of its contents to third parties is subject to the advance written consent of Kroll. Kroll assumes no
direct, indirect or consequential liability to any third party or any other person who is not the intended
addressee of this report for the information contained herein, its interpretation or applications, or for
omissions, or for reliance by any such third party or other person thereon. To the extent information provided
in this report is based on a review of publicly-available records, such information, as presented, relies upon
the accuracy and completeness of those records, which have not been corroborated by Kroll. Statements
herein concerning financial, regulatory or legal matters should be understood to be general observations
based solely on Kroll’s experience as risk consultants and may not be relied upon as financial, regulatory or
legal advice, which Kroll is not authorized to provide. All such matters should be reviewed with appropriately
qualified advisors in these areas.

Private and Confidential 2


Project Tenor II report

Contents
Introduction ........................................................................................................................ 7

1.1 Background ........................................................................................................................................... 7


1.2 Objective and scope .............................................................................................................................. 7
1.3 Reporting structure and distribution ....................................................................................................... 9
1.4 Contents of the Detailed Report ............................................................................................................. 9
1.5 Caveat ................................................................................................................................................. 10

Executive Summary ........................................................................................................ 11

2.1 Total loan analysis and destination ...................................................................................................... 12


2.2 Events which allowed such a significant increase in lending to take place – November 2014 .............. 12

2.2.1 Concentrating the loan exposure in BEM to increase liquidity ......................................................... 13

2.2.2 Questionable collateral for the November 2014 loans ..................................................................... 13

2.2.3 Controlling the loan approval process within the Three Moldovan Banks ........................................ 14

2.2.4 Switching the loan portfolio to BS using fake overdrafts.................................................................. 14


2.3 Shor Group ownership of the Three Moldovan Banks .......................................................................... 14
2.4 Shor Group control of the loan application process and accounts involved in the fraud ........................ 15
2.5 Laundering mechanisms...................................................................................................................... 15
2.6 Significant fund flows traced to individuals ........................................................................................... 17
2.7 Other fund flows by jurisdiction ............................................................................................................ 17

Methodology .................................................................................................................... 19

3.1 Statement of Mr. Shor.......................................................................................................................... 20


3.2 Caveat ................................................................................................................................................. 20
3.3 Exchange rates.................................................................................................................................... 20
3.4 The review period ................................................................................................................................ 21
3.5 The tracing methodology ..................................................................................................................... 21

Increase in the loan exposure 2011 – 2014 .................................................................... 22

4.1 Overview of exposures ........................................................................................................................ 22

4.1.1 Shor Group companies................................................................................................................... 22

4.1.2 Loan exposures – 2012 to 2014 ..................................................................................................... 23

4.1.3 Analysis of new loans issued – January 2012 to November 2014 ................................................... 24

4.1.4 Quantification of the loss in the Three Moldovan Banks.................................................................. 26


4.2 Events which allowed a significant increase in lending to occur – November 2014 .............................. 27

4.2.1 Concentration of new loans in BEM during November 2014 ........................................................... 27

4.2.2 Questionable Russian bank deposits in BEM ................................................................................. 27

4.2.3 Coordinated management of liquidity between the Three Moldovan Banks .................................... 29

4.2.4 Loan documentation and approval.................................................................................................. 30

Private and Confidential 3


Project Tenor II report

4.2.5 Coordinated loan applications at the Three Moldovan Banks.......................................................... 32


4.3 Switch of the entire loan portfolio on 25-26 November from BEM to BS ............................................... 33

4.3.1 Fake overdraft created in BS .......................................................................................................... 35

4.3.2 Reallocation of interbank deposits .................................................................................................. 35

Shor Group connections with the ownership of the Three Moldovan Banks .................. 36

5.1 Unibank (UB) ....................................................................................................................................... 37


5.2 Banca Sociala (BS) ............................................................................................................................. 38
5.3 Banca de Economii .............................................................................................................................. 39

Laundering of loan funds ................................................................................................. 41

6.1 Initial destinations of loan funds ........................................................................................................... 42

6.1.1 Funds to the Core Laundering Mechanism ..................................................................................... 43

6.1.2 Destination of funds that did not pass through the core laundering mechanism .............................. 44

6.2 Laundering mechanisms...................................................................................................................... 45

6.2.1 Core Laundering Mechanism.......................................................................................................... 45

6.2.2 Fund flows involving accounts at Russian banks ............................................................................ 50

6.2.3 Overview of amounts dissipated to Russia ..................................................................................... 52

6.2.4 Other laundering mechanisms identified ......................................................................................... 52

End destination of fraudulent loan funds ......................................................................... 54

7.1.1 Overview of end jurisdictions .......................................................................................................... 55

7.1.2 Ends in accounts in Latvia and Estonia .......................................................................................... 56

Identification of beneficiaries / perpetrators .................................................................... 57

Private and Confidential 4


Project Tenor II report

List of Tables
Table 1 : USD Exchange rates used in this report.................................................................. 21
Table 2 : Total loan exposure at the Three Moldovan Banks 2011-2014 ................................ 23
Table 3 : Ageing of loan exposure at 31 October 2014........................................................... 24
Table 4: Estimation of the total loss ....................................................................................... 26
Table 5 : Liquidity as reported to NBM in the daily balance sheets ...................................... 28
Table 6 : Initial destination of loan funds 2012 – November 2014 ......................................... 43
Table 7 : Summary of all funds received into the Core Laundering Accounts...................... 49
Table 8: Other Receipts traced out ......................................................................................... 50
Table 9: Net effect of transactions with Russian banks: ....................................................... 52
Table 10: overview of total end destinations of funds ........................................................... 54
Table 11: Amounts traced to ends by jurisdiction .................................................................. 55

Private and Confidential 5


Project Tenor II report

List of Figures
Figure 1 – New loans issued from 2012 to the end of November 2014 ....................................... 25
Figure 2 – Shor Group loan portfolio 1 - 24 November 2014 ...................................................... 25
Figure 3: Interbank deposits increase liquidity in BEM and UB/BS ............................................. 30
Figure 4: Details of fund flows to switch the portfolio from BEM to BS on 25 - 26 November 2014
.................................................................................................................................................. 34
Figure 5: Illustration of the funding of the purchase of UB shares ............................................... 38
Figure 6: Initial dissipation of loan funds .................................................................................... 42
Figure 7: Overview of core laundering mechanisms in Latvia ..................................................... 47

Private and Confidential 6


Project Tenor II report

Introduction
1.1 Background

Kroll’s Scoping Phase investigation (the “Scoping Phase”), as reported on 2 April 2015, found
evidence indicative of wrongdoing in relation to actions taken within Banca de Economii SA
(“BEM”), Banca Sociala SA (“BS”), and Unibank SA (“UB”), (together, the “Three Moldovan Banks”)
between 2012 and 2014, which appears to have ultimately resulted in their collapse at the end of
November 2014 with a combined loan exposure in the region of USD 1 billion.

The Scoping Phase identified that during the period 17 August 2012 to 30 November 2014, the
Three Moldovan Banks were consecutively subjected to significant shareholder changes, which
had the effect of transferring ownership to a series of apparently unconnected individuals and
entities. Thereafter, each of the banks entered into a series of questionable transactions, which do
not appear to have a sound economic rationale, and which ultimately resulted in such a significant
deterioration in each of their balance sheets that they were no longer viable as going concerns.

It was identified in the Scoping Phase that the banks appeared to have co-ordinated to maximise
available liquidity, in order to facilitate a massive increase in lending by the Three Moldovan Banks
to Moldovan entities. Preliminary analysis showed that these entities effectively formed a group of
related parties, whose loan receipts from the banks were passed through a complex web of
transactions using predominantly UK Limited Partnerships with Latvian bank accounts. Following
this complex laundering process, the majority of the loan funds were channelled back to Moldova
to pay down existing loan exposures, to create the impression that these were genuine corporate
loans, and to allow the related party loan portfolio to keep increasing, whilst approximately USD
600 million was dissipated to bank accounts across numerous jurisdictions.

Kroll’s Scoping Phase report found evidence that suggested that Ilan Shor, and companies and
individuals affiliated with him (the “Shor Group”) played an integral role in coordinating these
activities, suggesting that he was one of, if not the only beneficiary. The transactions to which these
Shor Group entities were party to appear to have ultimately contributed to the collapse of the banks
at the end of 2014.

1.2 Objective and scope


The overall objective of the investigation in Phase II was to evidence what caused the Three
Moldovan Banks to collapse, to understand who perpetrated and benefitted from the fraud, and to
support criminal and/or civil proceedings to recover the misappropriated funds.

The investigation has focused on understanding key transactions that took place in 2014,
particularly in BEM, during which time a significant proportion of the loss occurred, in order to clarify
the mechanisms used to increase lending and to identify documentary evidence of wrongdoing. In

Private and Confidential 7


Project Tenor II report

addition, the investigation analysed the loan portfolios of the Three Moldovan Banks from 1 January
2012 to the end of 2014, with the view of clarifying the timing of losses that occurred in this period
and identifying the ultimate destination of the funds originating from the fraudulent loans issued
and the beneficiaries of the wrongdoing.

NBM has, with the assistance of Kroll, developed a collaborative relationship with a number of
international regulators, particularly in Latvia and Estonia, which has allowed the investigation to
date to progress without formal legal disclosure applications. A Memorandum of Understanding has
also been signed between the NBM and the Central Bank of Cyprus (“CBC”), although disclosure
is awaited from CBC. It is likely that formal legal disclosure applications in a number of jurisdictions
will be required to progress the investigation, in order to identify the end destination of funds and
to maximise the chance of recovering stolen assets.

Kroll also identified accounts at other Moldovan banks which were used to further disguise fund
flows relating to the fraud, but to date has not identified any evidence of involvement or complicity
of the management or owners of these banks in the fraud. Further analysis and disclosure will be
required to clarify the nature of these transactions and the end destination of funds.

Broadly, the investigation in Phase II has focussed on answering the following questions:

• When did the loss of the USD 1 billion occur during the period 2012 to November 2014?

• How were BEM, BS and UB able to increase the corporate lending to Shor Group entities
to approximately USD 1 billion by the end of November 2014?

• How was the entire loan portfolio at BEM, of close to USD 1 billion switched from BEM to
BS on 25 and 26 November 2014?

• Who coordinated and controlled the transactions that enabled the increase in corporate
lending and the switching of the loan portfolio to take place?

• What mechanisms were used to launder the funds obtained through the fraudulent loans
between 2012 and 2014?

• How much of these new loans issued between 2012 and 2014 were used to pay off existing
loan exposure at the Three Moldovan Banks and/or other Moldovan banks?

• What was the final destination of funds that were not recycled to pay off existing loan
exposure?

• Who were the ultimate beneficiaries of the stolen funds?

• What are the priorities to maximise chances of recovery of assets?

Private and Confidential 8


Project Tenor II report

The next phase of our work will comprise taking steps to recover the misappropriated funds, as
well as the identification of parties who knowingly participated in and benefitted from the fraud.

1.3 Reporting structure and distribution

Following a meeting with representatives of the Moldovan Government, the NBM and investigative
and prosecutorial authorities in November 2017 (the “November Meeting”), Kroll and Steptoe and
Johnson (together the “Consortium”) were instructed to produce an additional report of investigative
work to date (the “Summary Report”), that will be made public, accompanied by a more detailed
report (the “Detailed Report”) containing specific details and examples of fund flows and
mechanisms engaged which allowed the fraud to occur, that will remain subject to initial
confidentiality undertakings and could be provided to relevant authorities, provided that those
authorities will provide first to Kroll and Steptoe and Johnson a written undertaking to maintain
confidentiality of the Detailed Report. In order to protect the potential for future recoveries and to
avoid prejudice of any ongoing criminal or other investigation in Moldova or elsewhere, it was
agreed at the November Meeting that information relating to the names of entities and individuals
that appeared to have benefitted from the proceeds of fraud would only be disclosed to the
Moldovan criminal authorities.

1.4 Contents of the Detailed Report


This document is a high level Summary Report which will be accompanied by additional details
included in the Detailed Report. The sections included in the Detailed Report are shown below:

Section Title of section


Section 1 Glossary of definitions
Section 2 Shor Group companies
Section 3 Process of approval of loans at BEM
Section 4 Details of loans granted by the Three Moldovan Banks
Section 5 Detailed fund flows of the Switch of funds from BEM to BS in
November 2014
Section 6 Fund flow analysis of Shareholding of the Three Moldovan Banks
Section 7 The Core Laundering Mechanism
Section 8 The forensic funds flow methodology
Section 9 Secondary laundering mechanisms identified
Section 10 The involvement of accounts at Russian Banks
Section 11 Detailed analysis of the destination of funds
Section 12 Inventory of electronic data collected and reviewed

Private and Confidential 9


Project Tenor II report

1.5 Caveat
In preparing this report, Kroll has placed reliance on documentation provided by NBM, as well as
overseas regulators, to draw a number of the conclusions set out in this report. Kroll makes no
representation as to the authenticity or completeness of the documentation provided by the NBM
or the regulators. There may be additional documentation or information to which Kroll has not had
access which could contradict or challenge conclusions drawn from the documentation reviewed to
date.

Analysis was conducted on documentation and transactional information available to date within
the Three Moldovan Banks. The data set extracted for review was based on specific targeted
searches. Additionally, meetings conducted to date were for the purposes of furthering a general
understanding of the activity in the Three Moldovan Banks, and no formal statements have been
retained and signed by the individuals interviewed at this stage. Kroll has not interviewed or
requested explanations from all individuals involved in the matters set out in this report.

Private and Confidential 10


Project Tenor II report

Executive Summary
Our investigation to date has identified contemporaneous and independent documentary evidence
that indicates that the Three Moldovan Banks were subjected to a large, coordinated fraud, which
took place over at least three years, and intensified in 2014, ultimately resulting in their collapse.
The suspected fraud involved the issuing of hundreds of loans to linked companies, the majority of
which were transferred to a laundering mechanism in Latvia. While the majority of the loan funds
were then channelled back to Moldova to repay existing loans and to allow the continuation of
lending, at least USD 600 million was dissipated to other destinations.

Contemporaneous documents suggest the involvement in the suspected fraud of a large group of
Moldovan companies working in concert, linked to Mr. Ilan Shor (“Mr. Shor”) (the “Shor Group”). At
least 77 companies with accounts at the Three Moldovan Banks made up this group. Individuals
and companies linked to the Shor Group increased their ownership interest in the Three Moldovan
Banks in an apparent attempt to facilitate the fraud, and thereby controlled the banks and caused
them to increase to a huge extent loans made by the banks. Companies linked to the Shor Group
also provided the corporate and bank account structures through which the fraud appears to have
been executed. Mechanisms identified which appears to have allowed the fraud to occur and
remain undetected included: misrepresentation of liquidity ratios by the banks to the NBM; the
securing of loans against questionable collateral; the exclusion of certain board members from
decision making processes and the ignoring of concerns raised by senior employees at the banks,
which together allowed large scale lending to continue to unsustainable levels in 2014.

Loan funds provided by the Three Moldovan Banks to Shor Group companies totalled USD 2.9
billion between 2012 and 2014. Loan funds provided to Shor Group companies, under the guise of
corporate lending, appear to have been laundered in a highly coordinated process through a
network of connected companies and bank accounts in Latvia, apparently also coordinated by Mr
Shor, before being dissipated to multiple bank accounts across numerous jurisdictions. As such,
these do not seem to be genuine commercial loans at all but were used as fronts to fraudulently
misappropriate millions of dollars from the Three Moldovan Banks.

This report analyses the destination of USD 2.9 billion of loans which were issued by the Three
Moldovan Banks between 2012 and 2014. The majority of these loan funds were channelled
through what appears to be a coordinated money laundering mechanism in Latvia involving two
Latvian banks (the “Core Laundering Mechanism”). Following the apparent laundering of funds
through the Core Laundering Mechanism, the funds were either returned to Moldova to repay other
loans and thereby allow lending to continue, or were subjected to further money laundering
mechanisms involving other Moldovan, Latvian and Estonian banks, before being dissipated to
accounts in multiple other jurisdictions.

Private and Confidential 11


Project Tenor II report

2.1 Total loan analysis and destination

The total loan exposure of the Three Moldovan Banks increased from USD 491 million at the
beginning of 2012 to approximately USD 1 billion by the end of November 2014. By the end of
October 2014, nearly 70% of the total loan exposure was to Shor Group companies. This exposure
increased again during November 2014 to around 80% of the total loan assets of the Three
Moldovan Banks.

The majority of the loan proceeds were used to pay down existing loans, after being apparently
laundered through the Core Laundering Mechanism, while a significant proportion of the funds was
extracted from the Core Laundering Mechanism, and appear to comprise stolen assets. The total
amount dissipated was spread throughout the period, with subsequent new loans being issued in
an attempt to disguise the previous dissipation and to pay off the loans from which funds were
misappropriated. Analysis of the initial destination of loan funds revealed that in the period January
2012 to April 2014, the majority of the funds (totalling USD 478 million) were transferred to bank
accounts at the first of the Latvian banks involved in the Core Laundering Mechanism (“Latvian
Bank 1”). From May to November 2014, a second Latvian bank (“Latvian Bank 2”) became the main
vehicle for laundering the loan funds totalling USD 2.1 billion.

Of the USD 2.9 billion, approximately USD 220 million remained in Moldova and was used to repay
loans at the Three Moldovan Banks as well as at other Moldovan banks. Approximately USD 110
million of loan funds was transferred directly accounts in other jurisdictions, including to accounts
held at Russian banks, without passing through the Core Laundering Mechanism. It appears that
at least part of this was an additional laundering mechanism to add a further layer of opacity to the
fund flows. As detailed further in section 2.5, approximately USD 2.6 billion was passed to the Core
Laundering Mechanism, and approximately USD 2.0 billion was returned back to accounts at the
Three Moldovan Banks. This left a value of approximately USD 600 million which appears to have
been stolen following the Core Laundering Mechanism and dissipated to other destinations.

2.2 Events which allowed such a significant increase in


lending to take place – November 2014

As noted in the Scoping Phase report, the majority of documentation relating to the loans issued to
Shor Group companies was destroyed in suspicious circumstances just prior to the collapse of the
Three Moldovan Banks. Kroll has analysed daily balance sheets reported to the NBM and electronic
data contained on the computers and servers from the Three Moldovan Banks, in order to
understand the internal processes that allowed such a significant increase in lending, particularly
focussing on November 2014, during which the increase was most prominent. The key findings
from this analysis are as follows:

Private and Confidential 12


Project Tenor II report

1. First, during the period 1 to 24 November 2014, the existing loan exposure to the Shor
Group, which was spread across all Three Moldovan Banks, was concentrated in
BEM. This allowed the Three Moldovan Banks to pool their liquidity into BEM and enabled
BEM to increase lending.

2. Second, documents strongly suggest that collateral used to secure the increasing loan
portfolio, namely cash placements at Russian banks made as a result of interbank deposits
from the same Russian banks, did not exist. It appears that such collateral was recorded
on the books of the Three Moldovan Banks to provide an illusion of liquidity to the NBM and
to allow further lending.

3. Third, credit committee approval for loans was bypassed and loans were only approved by
the boards of the Three Moldovan Banks. Attendance at Board meetings was controlled
and certain members were not invited to attend the meetings at which the November 2014
loans and related transactions were approved.

4. Finally, the entire loan portfolio was switched from BEM to BS on 25 and 26 November
2014 by creating a fake overdraft facility with Latvian Bank 2 in the books of BS. The
purpose of this switch of the loan portfolio from BEM to BS is not yet clear, but had the
effect of concentrating the entire loan portfolio into BS and allowing a further
misappropriation of funds to take place due to the additional liquidity created by the switch.

2.2.1 Concentrating the loan exposure in BEM to increase liquidity

During November 2014, the majority of new loans were issued by BEM, while the exposure at UB
and BS reduced. Existing loans at UB and BS were repaid from the newly issued loans from BEM,
with the majority of funds passing through the Core Laundering Mechanism in an apparent attempt
to disguise the origin of funds. This had the effect of consolidating all the loans from the Three
Moldovan Banks at BEM by 24 November 2014.

2.2.2 Questionable collateral for the November 2014 loans

Throughout November 2014, there was a coordinated management of liquidity between the Three
Moldovan Banks, with interbank deposits being used to increase liquidity and fund new loans. In
addition, Russian bank deposits were recorded in the daily balance sheets reported to NBM as
providing collateral for the new loans issued by BEM.

Documents suggest that the Russian banks did not provide deposits as collateral for the new
November loans and that these deposits were falsely created in the books of BEM to facilitate the
increase in lending and to present an illusion of liquidity to the NBM.

Private and Confidential 13


Project Tenor II report

2.2.3 Controlling the loan approval process within the Three Moldovan Banks

Loans were approved by the Boards of Directors, despite warnings from senior management that
they appeared to be to related parties who could not demonstrate the business purpose or the
financial stability to repay the loan. Employees did not challenge the Board or escalate to any other
authority. Kroll (through NBM) contacted two board members appointed by the Moldovan
government as part of the supervision of BEM, who stated that they had not been invited to any
Board meetings in November, and were not aware of any loans being granted during this period.

It is not currently clear how much involvement other members of the Board had, although a review
of minutes of Board meetings recovered from the electronic data set indicates that all other
members were present at the meetings where the loans and other related transactions were
approved. From interviews with staff members and a review of the limited documentation available,
it is apparent that in numerous instances, risks were raised to the Board but appeared to have been
ignored and the loans were approved.

2.2.4 Switching the loan portfolio to BS using fake overdrafts

Analysis of the customer loan account statements within the Three Moldovan Banks and the Latvian
accounts has confirmed that on 25 and 26 of November 2014, the entire corporate loan portfolio
was switched from BEM to BS. The methods used to implement the switch included the creation of
a fake overdraft of BS on its correspondent account with Latvian Bank 2, to provide liquidity for new
loans to be issued, the reallocation of interbank deposits between BEM and BS, and the return
(deletion) of the Russian bank deposits from BEM’s balance sheet. Following this switch, BEM’s
only assets were the interbank deposits held at BS.

A senior employee at BS confirmed that they had been instructed by senior management at BS to
process the fake overdraft with Latvian Bank 2, which had the effect of increasing liquidity to allow
the switch of the loan portfolio to take place.

2.3 Shor Group ownership of the Three Moldovan Banks

Kroll has traced the funds used to acquire shares in the Three Moldovan Banks and has identified
that the origin of the funds was largely funded by Shor Group companies, either in part from loans
granted by the Three Moldovan Banks, or from what appear to be loans to companies or individuals
linked to Shor with accounts at two Russian banks. The analysis showed that the Core Laundering
Mechanism was used to apparently launder these funds, which had the effect of disguising the
source of funds and giving the false impression that the shareholders were unrelated.

This demonstrates that there appears to have been a concerted effort on the part of the Shor Group
to gain control of the banks as early as 2012 / 2013 and to conceal this through the use of an
established network of nominees, shell companies and offshore accounts. The acquisition of

Private and Confidential 14


Project Tenor II report

significant stakes in the banks ultimately facilitated the Shor Group’s ability to exercise control over
them and to facilitate the suspected fraud.

2.4 Shor Group control of the loan application process and


accounts involved in the fraud

As well as the ownership of the banks and the establishment of a series of linked companies that
worked in concert to obtain fraudulent loans, review of email data recovered from the Three
Moldovan Banks demonstrated that the Shor Group centrally controlled and coordinated loan
applications within the Three Moldovan Banks and maintained close communication with senior
employees at the Banks.

In addition, account opening documentation and other correspondence has confirmed that the
accounts at Latvian Bank 2 were affiliated to and/or controlled by the Shor Group and that Mr. Shor
dealt directly with Latvian Bank 2 and he provided purported explanations for the purpose of the
multiple accounts and the millions of dollars transiting through them. Disclosure has not yet been
obtained from Latvian Bank 1. However, accounts in the names of both Mr Shor and other
individuals affiliated with him received funds originating from the fraud into accounts held at Latvian
Bank 1, suggesting a similar mechanism. Furthermore, according to bank account opening
documentation, Mr. Shor was listed as the UBO for one of the companies that was used to launder
funds through its account at Latvian Bank 1.

2.5 Laundering mechanisms

As noted above, the Core Laundering Mechanism involved multiple bank accounts at two Latvian
banks. In total, Kroll analysed fund flows through the Core Laundering Mechanism totalling USD
2.6 billion between 2012 and 2014. A total of 81 bank accounts at the two Latvian banks were
identified as “Core Laundering Accounts”, which appeared to have been established solely for the
purpose of laundering the funds, as they did not have any other significant or ordinary commercial
transactions. The majority of these accounts were held in the names of UK limited partnerships, or
companies registered in offshore locations such as Belize, BVI and Panama. Most of these
companies had recently been incorporated, they had no established business, no public accounts,
no business premises, no other bank accounts, no public profile and no identifiable owners.

The laundering mechanisms included the frequent transfer of funds between linked accounts, the
splitting of funds and layering through other accounts, simultaneous issuing and repaying of an
overdraft by two linked companies to disguise the onward flow and the frequent and apparently
arbitrary switching of currencies between accounts. The mechanism was a highly coordinated
professional laundering operation. The complexity and timing of the transactions indicates the use
of a complex algorithm that had the ability to execute a high volume of transactions through internet

Private and Confidential 15


Project Tenor II report

banking portals within a very short period of time. There are numerous examples where funds
flowed through a series of accounts within seconds.

In addition to the Core Laundering Mechanism, Kroll identified other peripheral laundering
mechanisms that appear to be another part of the money laundering operation. First, Kroll traced
transactions exiting the Core Laundering Mechanism to Estonia totalling USD 58 million. Further
analysis of the onward transfers of funds in Estonia identified a further 45 bank accounts held by
UK partnerships or companies in offshore jurisdictions that appeared to be part of a larger
laundering mechanism, as they had high volumes of activity and similar transaction patterns as the
Latvian accounts, for example multiple splitting of values, apparently arbitrary currency changes,
and vague and inconsistent descriptions of goods purportedly being acquired using the loan
proceeds. Analysis of the account opening documentation of the companies and IP addresses from
which the internet banking transactions were processed identified links between many of the
Estonian companies, as well as some commonality with accounts in the Latvian laundering
mechanisms.

In addition to the laundering mechanisms in Latvia and Estonia, laundering mechanisms were
identified involving accounts at other Moldovan banks. Examples were identified where funds were
moved through multiple company accounts at other Moldovan banks, either before being
transferred to the Core Laundering Mechanism, or following the Core Laundering Mechanism, to
further conceal onward fund flows. From documentation available to date, no evidence has been
identified which indicated complicity of the management or ownership of these banks in the fraud.

Furthermore, it appears that accounts at Russian banks were also used in a further mechanism to
blur the origin and destination of funds. First, accounts in the Core Laundering Mechanism received
funds amounting to approximately USD 200 million from accounts held at two Russian banks.
Analysis of the amounts and the source of the funds indicate that the origin of funds was from a
series of loans that were repaid at a later date from loans made by the Three Moldovan Banks to
Shor Group companies. The accounts held at the Russian banks were also in the name of Shor
Group companies, individuals who were the directors of Shor Group companies, or other Moldovan
individuals.

In a second scheme, USD 100 million was received on accounts at BEM from a third Russian bank,
and was placed on overnight deposit, according to BEM’s records and the information provided to
the NBM. However, from a review of the bank statements of BEM’s shareholders who received a
proportion of these funds, it appears that these funds were actually transferred to the Core
Laundering Mechanism, and were not placed back on deposit. According to SWIFT messages,
these funds were returned a year later, but were reported on the balance sheet of BEM for almost
two further months, resulting in an overstatement of BEM’s liquidity ratios in September and
October 2014.

Private and Confidential 16


Project Tenor II report

Analysis of fund flows also revealed possible further laundering schemes that have yet to be
analysed in detail due to current lack of documents. The first involved significant transfers to bank
accounts in China and Hong Kong, purportedly for the purchase of textiles and construction
materials. It has not been possible to corroborate whether these transactions relate to bona fide
trade transactions, although there were frequent occurrences when the description of the goods
apparently being purchased was changed as the transaction continued through an earlier
laundering mechanism, which casts doubt on the veracity of the description of the end transaction.
Additionally, a significant proportion of the funds went to Cyprus. As some of the same companies
were found to have transferred funds back to other Latvian accounts, it appears likely that this is
another part of an organised laundering mechanism. Requests have been made to the CBC for
assistance and disclosure, but no disclosure has yet been provided.

2.6 Significant fund flows traced to individuals

The investigation has identified several Moldovan-based individuals who appear to have received
benefit from or are connected to the fraud within the Three Moldovan Banks. Receipts into personal
accounts or into companies known to be linked to these individuals totalled approximately USD 50
million. In order to avoid prejudice to any current or future legal proceedings, this information will
be shared only with the appropriate criminal authorities in Moldova.

2.7 Other fund flows by jurisdiction

Of the USD 2.6 billion which was transferred to the Core Laundering Scheme detailed above, USD
2.0 billion was returned to Moldova following laundering of funds through multiple linked company
accounts. Kroll has traced the remaining USD 600 million out of the Core Laundering Mechanism
onwards, subject to the availability of disclosure of bank statements and cooperation with
authorities in the jurisdictions identified. The other fund flows that did not initially go to the Core
Laundering Mechanism were also traced where possible (totalling approximately USD 300 million).
These funds were predominantly used to repay other loans, but there is a proportion of these which
could also have resulted in further loss. Therefore the total loss identified was at least USD 600
million and potentially as much as USD 900 million.

Given the mixing of funds, layering and other money laundering schemes, the methodology used
for tracing funds involved tracing whole amounts if any onward payments that originated from the
fraud were included in those amounts. This means that the final amount of funds that have been
traced to end destinations is significantly larger than the USD 600 million that has been identified
as leaving the Core Laundering Mechanism. Kroll has identified end transactions totalling
approximately USD 1.0 billion of funds which originated at least in part from the suspected
fraudulent loans.

Private and Confidential 17


Project Tenor II report

The destination of onward payments by jurisdiction totalling USD 1.0 billion funded in whole or in
part by the fraudulent loans can be summarised as follows:

• Latvia – USD 302.1 million


• Moldova –other Moldovan banks (USD 169.3 million)
• Cyprus – USD 112.4 million
• China – USD 83.3 million
• Russia – USD 80.6 million
• Austria – USD 56.1 million
• Estonia – USD 51.4 million
• Switzerland – USD 42.8 million
• USA – USD 25.5 million
• Hong Kong – USD 22.6 million
• Other (multiple jurisdictions) – USD 115.5 million as follows:
o USD 54.1 million to accounts in jurisdictions that received less than USD 20 million
in aggregate;
o USD 50.5 million to jurisdictions that received less than USD 10 million in
aggregate; and,
o USD 10.9 million to accounts in jurisdictions that received less than USD 1 million.

In addition to the USD 1.0 billion detailed above, a further USD 82.9 million was traced to accounts
at the Three Moldovan Banks to repay loans.

As disclosure of the bank accounts for the ends currently identified has not yet been obtained, the
final destination of the funds which flowed to these accounts is currently unknown. It is likely that
the funds were transferred through these accounts to other destinations or to purchase assets. It
will be necessary to receive the cooperation of the relevant supervisory authorities or obtain
disclosure orders to further trace the dissipation of the stolen funds through these and other
jurisdictions, and to identify beneficiaries.

Private and Confidential 18


Project Tenor II report

Methodology

In order to address the questions outlined in the introduction, Kroll has undertaken the following
areas of work:

• Conducted a high level analysis of the increase in loan exposure of the Three Moldovan
Banks between 2012 and 2014;
• Analysed in detail the new loans issued to Shor Group companies in the period 2012 to 2014
to identify the initial destination of loan funds (i.e. whether they were transferred out of or
remained in Moldova);
• Conducted fact finding interviews with employees of the Three Moldovan Banks, to facilitate
document requests and to progress understanding of the control structures and processes
in place;
• Identified and analysed the mechanisms which allowed the loan funds to be laundered;
• Traced the loan funds to identify the end destination of fraudulent loan funds following the
laundering of funds through the Core Laundering Mechanism. This has involved the review
of tens of thousands of transactions, involving hundreds of companies and accounts in
multiple jurisdictions.
• Established and maintained cooperation with regulators in Latvia and Estonia to obtain
relevant disclosure and to facilitate exchange of information between NBM and these
authorities (and met with representatives of those regulators);
• Facilitated the signing of a Memorandum of Understanding between NBM and the
supervisory authorities in Cyprus;
• Held a meeting with the supervisory authority in Russia with a view of establishing a
mechanism for sharing of information between the Central Bank of Russia and the NBM;
• Liaised with international bodies to assist investigations into the fraud and the laundering /
dissipation of funds, including the EU, IMF, World Bank and US State Department;
• Held a meeting with senior management of a Latvian bank and their lawyers;
• Held a number of meetings with the anti-corruption prosecutor in Moldova in collecting and
evidencing data relating to criminal cases which have been opened;
• Identified and analysed the involvement of individuals responsible for both the administration
and implementation of the suspected fraud and those who appear to be beneficiaries of the
suspected fraud; and
• Created an inventory of electronic data that has been secured to date from the Three
Moldovan Banks, and secured further data, where available and relevant and identified
electronic data which may be relevant, uploaded this to an electronic review platform, and
performed a targeted search of documentation based on key search terms. Kroll selected for
review the electronic email data of key members of staff at the Three Moldovan Banks who
would most likely have knowledge and visibility of the loan issuing and approval process,

Private and Confidential 19


Project Tenor II report

such as board members, senior members of the credit and treasury departments and other
senior managers.

3.1 Statement of Mr. Shor

Kroll was made aware of the existence of a statement that has been uploaded onto the internet,
which purports to be Mr. Shor’s statement to the Public Prosecutor, providing an explanation for
some of the events that took place within the Three Moldovan Banks. Kroll has not been provided
with an official copy of Mr. Shor’s statement and has not had the opportunity to meet with Mr. Shor.
As such we are not at this stage in a position to confirm the authenticity of the document and do
not comment on its content within this report.

3.2 Caveat

In conducting this investigation, Kroll has relied upon information and analysis conducted by NBM,
supplemented by analysis of source data provided by external regulatory authorities as well as from
open source and public record research. External, investigative research has been conducted into
relevant individuals and entities as the investigation has progressed in order to identify the extent
to which transactions might be associated and individuals linked, as well as the likely legitimacy of
beneficiaries of transactions.

Due to the significant complexity of the transactional web and the prevalence of coordinated money
laundering across multiple jurisdictions, it has not been possible during investigations to date to
trace the entirety of the loan funds that were dissipated, or to demonstrate links between every
organisation that received funds. Not all potentially relevant information available has been
analysed, which may lead to the identification of further relevant parties and transactions as the
investigation continues.

Kroll has not been provided with all documentation requested to date as multiple document
requests were made from various sources, not all of which have been fulfilled.

3.3 Exchange rates


Loans were issued by the Three Moldovan Banks in MDL, USD and EUR. As the funds were moved
through the complex laundering mechanisms, they were subject to multiple changes in currency.
For the purposes of comparison and for consistency, all amounts in this report are presented in
USD equivalent amounts. The average midpoint exchange rate for each year 1 was used to translate
the original currency amounts to USD. These rates were obtained from Bloomberg. 2

1
Rounded to two decimal places
2
Note: These foreign exchange rates have been applied to all workings referred to in this report. In some cases, USD
equivalent amounts referred to in previous reports may have been updated with the foreign exchange rates in Table 1 for
the sake of consistency and accuracy.

Private and Confidential 20


Project Tenor II report

Table 1 : USD Exchange rates used in this report

EUR MDL RUB RON GBP UAH CHF


2012 1.29 0.08 0.03 0.29 1.59 0.12 1.07
2013 1.33 0.08 0.03 0.30 1.56 0.12 1.08
2014 1.33 0.07 0.03 0.30 1.65 0.08 1.09

3.4 The review period


The main focus of this investigation was to develop a detailed understanding of the transactions
and events that occurred following the issuing of loans by the Three Moldovan Banks between
1 January 2012 and 31 December 2014. (“The Review Period”). The Review Period was specifically
targeted as preliminary analysis by Kroll identified that it was during this period that the largest
increase in loans took place. A high level analysis of loan balances from January 2011 was also
undertaken.

3.5 The tracing methodology


A large proportion of the funds was transferred through a number of mechanisms involving banks
in Latvia, Estonia and Russia and movements of funds that appeared to have no commercial
purpose other than to disguise the origin of funds. As the fraud funds were transferred through
these laundering mechanisms, in a number of instances they were mixed with other funds that
could not be directly linked to the fraud. Where possible, these other funds have been separated
and excluded from the analysis. Where these other funds could not be separated, they were treated
as being tainted with suspected fraud funds and the combined transfers were traced for the
purposes of developing the asset recovery strategy.

As these mixed funds were treated as tainted with the suspected fraud funds, the total funds
identified at the end destination as a result of the further tracing work exceed the amounts originally
transferred to into the laundering mechanisms. Specifically, if a debit from an account was identified
as originating in whole or in part from suspected fraud funds, this amount was traced on in full. This
has had the result of identifying the end destination of fund flows that in large part contain funds
originating from the suspected fraud. As a result, the output amounts are significantly higher than
the input amounts, as the suspected fraud funds are split between multiple end amounts. Further
details of the methodology are provided in the Detailed Report.

Private and Confidential 21


Project Tenor II report

Increase in the loan exposure 2011 – 2014

The Scoping Phase report highlighted that the related party loan exposure of the Three Moldovan
Banks to Shor Group companies increased from USD 10.4 million 3 to USD 951 million 4 in the period
between September 2010 and 24 November 2014, which ultimately resulted in the collapse of the
Three Moldovan Banks.

A number of suspicious events relating to the extension of credit to the Shor Group and the
destruction of loan documents, were highlighted in the Scoping Phase report. In addition, on 25
and 26 November 2014, the loan portfolio was transferred from BEM to BS, before it was sold to a
UK registered limited partnership by the name of Fortuna LP.

Kroll has extended the review from the work conducted in the Scoping Phase to provide a more
detailed analysis of the exposure and loans issued during the Review Period. Further details are
provided in the section below.

4.1 Overview of exposures


The loan exposures of the Three Moldovan Banks to the Shor Group increased substantially from
the beginning of 2012 to 24 November 2014, the day before the banks collapsed. Between 1 and
24 November 2014, the portfolio of Shor Group loans was moved from UB and BS to be
concentrated in BEM. On 25 to 26 November 2014, the entire loan portfolio was switched from
BEM to BS. These activities demonstrate that there was a sophisticated co-ordination of control of
the Shor Group’s loan portfolios with the Three Moldovan Banks.

Kroll has analysed the loan books of the Three Moldovan Banks from 1 January 2012 to 24
November 2014 as preliminary analysis identified that the most significant increases in exposure
to the Shor Group occurred during this period. (This analysis is detailed further in section 4.1.3.)

4.1.1 Shor Group companies


The NBM previously identified 39 companies as being related to the Shor Group, as detailed in the
Scoping Phase report. In addition to these, further analysis by the NBM and Kroll has identified a
total of 77 companies linked to Shor. Of these, 50 Shor Group companies received loans from the
Three Moldovan Banks during the Review Period.

A list of the Shor Group companies, which appear to have worked in concert to fraudulently
embezzle money from the Three Moldovan Banks before transferring it to the Core Laundering
Mechanism, is included in the Detailed Report.

3
2012 rate used
4
The Scoping Phase report referred to an increase in MDL – here USD equivalents are given for consistency with the
remainder of the report.

Private and Confidential 22


Project Tenor II report

4.1.2 Loan exposures – 2012 to 2014


Exposure to the Shor Group increased from the beginning of 2012 to 2014, and the increase
accelerated during 2013 and the first 10 months of 2014. The value of the Shor Group exposure
increased by 70% between 1-24 November 2014, from USD 554 million to USD 951 million. During
this period, the proportion of the Shor Group exposure compared to the total exposure of the Three
Moldovan Banks increased from 25% at the beginning of 2012 to 80% at 24 November 2014.

Table 2 summarises the total loan exposure of the Three Moldovan Banks during this period and
details the increasing proportion of exposure to Shor Group companies:

Table 2 : Total loan exposure at the Three Moldovan Banks 2011-2014


Loans and BEM (USD BS (USD UB (USD TOTAL Of Which Shor
advances millions) millions) millions) Group 5
USD % of
millions total
31 December 2011 258 199 34 491 124 25%
31 December 2012 200 166 52 418 119 28%
31 December 2013 118 232 266 616 294 48%
31 October 2014 261 257 298 816 554 68%
24 November 2014 994 128 59 1,181 951 80%

The exposures shown in Table 2 which increased the loan exposure to Shor Group companies in
2013 and 2014 were consolidated in BEM in November 2014 through the issuing of new loans in
the period 1-24 November 2014. These loans were then transferred in their entirety from BEM to
BS on 25 and 26 November 2014, before being transferred to a UK limited partnership, Fortuna LP
(detailed further in section 4.3).

An analysis of the ageing of the loan balance 6 as at 31 October 2014, which was then further
increased in the month of November (across the Three Moldovan Banks) indicates that a significant
portion (85%) of the loan exposure at that date related to loans issued between 2012 and 2014.
Only 15% of the loan exposures related to loans issued before 2012. This indicates that the majority
of the loans issued pre-2012 were repaid by the end of October 2014. Analysis to date has not
identified significant transfers from Shor Group accounts to companies that held pre-2012 loans at
the Three Moldovan Banks, therefore it does not appear that the pre-2012 loans were paid off to a
significant level by Shor Group loans issued by the Three Moldovan Banks.

5
This includes the loans which were repaid during the “Roseau” cession agreement, whereby approximately USD 80
million of loans were repaid as part of the agreement for the change in ownership of BEM.
6
The ageing analysis was conducted on data at 31 October 2014 (instead of 24 November 2014) from the monthly
ageing of the loan book.

Private and Confidential 23


Project Tenor II report

Table 3 : Ageing of loan exposure at 31 October 2014


Loan issue BEM BS UB (USD Total (USD % of Oct 2014
date (USD (USD million) million) balance issued
million) million) in each year

Pre-2012 74 48 3 125 15%


2012 8 23 3 34 4%
2013 11 73 142 226 28%
2014 7 168 113 150 431 53%
Total 261 257 298 816 100%

4.1.3 Analysis of new loans issued – January 2012 to November 2014


To enable the Shor Group to pay down existing loans and to service the interest payments on those
loans, while allowing a proportion of funds to be dissipated to other destinations, the total of the
loans issued to the Shor Group had to increase during the Review Period. From January 2012 to
the end of November 2014, new loans were issued to Shor Group companies amounting to
approximately USD 2.9 billion. The majority of the loan proceeds were used to pay down existing
Shor Group loans, after being laundered through a complex network of bank accounts, while a
proportion of the funds was dissipated to other destinations. The total amount dissipated was
spread throughout the period, with subsequent new loans being issued to disguise the previous
dissipation and to pay off the loans from which funds were stolen.

Figure 1 shows the value of new loans issued by the Three Moldovan Banks between 1 January
2012 and 24 November 2014 to the 50 Shor Group companies that received loans (see section
4.1.1). In the twelve months ending December 2012, USD 136 million in new loans were issued to
these companies. This amount increased during 2013 (USD 373 million in new loans) and the first
10 months of 2014 (USD 504 million in new loans). From 1 to 24 November, a total of USD 1.0
billion in new loans was issued to these companies, mainly by BEM and BS.

On the 25 and 26 November 2014, the entire loan portfolio was moved from BEM to BS through
the issuing of a further USD 1.0 billion of loans by BS and the paying down of all existing loans at
BEM. Further details of the process by which this was implemented can be found in section 4.3.

7
1 January to 31 October 2014

Private and Confidential 24


Project Tenor II report

Figure 1 – New loans issued from 2012 to the end of November 2014

New loans issued over time (USD)


2,500,000,000

2,000,000,000

1,500,000,000

1,000,000,000

500,000,000

-
2012 (Jan to Dec) 2013 (Jan to Dec) 2014 (Jan to Oct) 2014 (Nov)

BEM BS UB Total

Figure 2 shows the lending by BEM to the Shor Group during the period 1 to 24 November 2014,
compared to the lending to the Shor Group by UB and BS during the same period. The figure shows
that although some loans were issued by BS and UB during this period, the loan balances at these
two banks was decreasing as the loans issued by BEM were used to repay existing loans at BS
and UB.

Figure 2 – Shor Group loan portfolio 1 - 24 November 2014

Shor Group loan exposure 1-24 November 2014


1,200,000

1,000,000

800,000
USD '000

600,000

400,000

200,000

-
02/11/2014
03/11/2014
04/11/2014
05/11/2014
06/11/2014
07/11/2014
08/11/2014
09/11/2014
10/11/2014
11/11/2014
12/11/2014
13/11/2014
14/11/2014
15/11/2014
16/11/2014
17/11/2014
18/11/2014
19/11/2014
20/11/2014
21/11/2014
22/11/2014
23/11/2014
24/11/2014

Date
BEM BS USD UB USD Total USD

Private and Confidential 25


Project Tenor II report

4.1.4 Quantification of the loss in the Three Moldovan Banks


Based on the analysis conducted to date, the loss suffered as a result of the suspected fraud at the
Three Moldovan Banks amounts to at least USD 600 million. The basis of this estimation is detailed
below. As additional analysis is conducted, particularly relating to accounts at other Moldovan
Banks and Latvian and Estonian accounts, this amount may be revised.

Table 4: Estimation of the total loss


Fund flow Amount Description Link to
(USD Report
billion) section
Loans issued 2.9 Loans issued by the Three Moldovan Banks 4.1.3
from 1 January 2012 to 26 November 2014
Loss calculated as follows:
Amounts transferred to 2.6 6.2.1.1
Core Laundering
Mechanism
Amounts returned from (2.0) Amounts returned to accounts at the Three 6.2.1.2
the Core Laundering Moldovan Banks from Core Laundering
Mechanism Accounts
Total loss: 0.6

Funds which were not transferred to the Core Laundering Mechanism totaling USD 0.3 billion have
been traced where possible. These funds were predominantly used to repay other loans at the
Three Moldovan Banks and other Moldovan Banks. Combined with the USD 600 million identified
above, the total loss could be as much as USD 900 million, taking this into account. Further
disclosure would be needed to clarify this.

Private and Confidential 26


Project Tenor II report

4.2 Events which allowed a significant increase in


lending to occur – November 2014
A series of events took place in the Three Moldovan Banks that allowed the extraordinary increase
in exposures to the Shor Group to occur in what appears to have been a concerted effort to strip
the Three Moldovan Banks of all liquidity before their inevitable collapse. In order to further
understand these events, Kroll focussed on the period between 1 and 24 November 2014 in BEM
as that was when the greatest increase in loans issued took place.

The Scoping Phase report highlighted that a large number of documents relating to the issuance
of loans at the Three Moldovan Banks disappeared in suspicious circumstances shortly prior to the
banks being placed under special administration on 28 November 2014. A van operated by a
company linked to the Shor Group used to transport BEM files from the bank for archiving was,
according to BEM staff, stolen and later found burned.

The imaging and processing of electronic data identified that individuals within the Three Moldovan
Banks had attempted to delete electronic files relating to the issuing of the loans and other internal
control information, or that documentation was not retained on local computers. However, Kroll was
able to identify fragments of the loan documentation through the forensic analysis of the custodian
computers and emails. This, combined with onsite interviews, has furthered our understanding of
the processes and methods behind the issuing of the loans and enabled us to determine the extent
to which management was aware of or involved in the large increase in loans during this period.

4.2.1 Concentration of new loans in BEM during November 2014


As shown in Figure 1 and Figure 2, the increase in lending from 1 – 24 November 2014 was mainly
concentrated in BEM, which issued USD 838.5 million in new loans during this period.

4.2.2 Questionable Russian bank deposits in BEM

According to the balance sheets that Kroll understands were submitted on a daily basis by BEM to
NBM during November 2014, the security that allowed BEM to issue new loans to Shor Group
companies in November 2014 was in the form of Russian bank deposits totalling USD 876 million.
According to the daily balance sheets, these bank deposits, received as collateral for the loans,
were simultaneously put on deposit at the same Russian bank, as cash on deposit. It appears that
this allowed BEM to create the illusion that it had sufficient liquidity to make loans while remaining
within the limits of the NBM liquidity requirements.

Liquidity requirements

In order to avoid excessive risk in the financial system and to protect the interest of depositors, the
NBM required that banks in Moldova maintained a level of liquidity, according to the following
regulation: the sum of a bank’s liquid assets must be not less than 20% of the sum of the total

Private and Confidential 27


Project Tenor II report

balance sheet assets. 8 Liquid assets are defined as “cash and precious metals, deposits with the
National Bank of Moldova, Liquid Securities and Current Net Interbank Resources”. Total assets
include longer term savings or placements with other banks, and other illiquid assets. This
assessment of liquidity does not take into account corresponding current liabilities.

Misrepresentation of liquidity to NBM

The Three Moldovan Banks submitted daily balance sheet summaries to NBM during November
2014, including a calculation demonstrating their current liquidity. A summary of these submissions
is provided in Table 5, both with and without the liquidity provided by the declared Russian bank
deposits. This highlights that without the additional cash asset in the form the deposits being placed
back as cash on account, the liquidity ratio would have fallen below the required 20% liquidity as
required by the NBM directive.

Table 5 : Liquidity as reported to NBM in the daily balance sheets

Liquidity as reported to NBM in the daily balance sheets – including liquidity provided by
the Russian bank deposits
Liquidity ratio
3-Nov-2014 12-Nov-2014 17-Nov-2014 18-Nov-2014 24-Nov-2014
reported by
BEM 62% 44% 40% 40% 30%
BS 36% 54% 50% 50% 54%
UB 22% 50% 62% 62% 60%
Combined 43% 46% 44% 44% 36%

Liquidity adjusted for position without the Russian bank deposits

Liquidity ratio 3-Nov-2014 12-Nov-2014 17-Nov-2014 18-Nov-2014 24-Nov-2014


adjusted for
BEM 56% 11% 1% 0% -30%
BS 18% 52% 48% 50% 54%
UB 0% 41% 57% 57% 60%
Combined 30% 25% 21% 20% -1%

This analysis demonstrates that without the cash on deposit at Russian banks, UB and BS were
already falling below the required levels of liquidity on 3 November 2014. The liquidity of BS and
UB improved towards the middle of November, largely due to additional injections of liquidity from
BEM, which was able to provide this predominantly because of the cash held at the Russian banks.
Without this additional liquidity, BEM would have been insolvent according to the NBM criteria by
17 November 2014. By 18 November 2014, it would have had no liquid assets.

8
http://www.bnm.org/en/content/regulation-banks-liquidity-approved-dca-nbm-no28-august-08-1997

Private and Confidential 28


Project Tenor II report

Lack of evidence to support Russian bank deposits


Despite the extent of the recorded Russian bank deposit activity during November 2014, Kroll’s
review of electronic and hard copy documentation obtained from the Three Moldovan Banks did
not identify any third party documentation which confirmed that deposits from the Russian banks
actually existed.

The nostro account 9 of BEM contained entries totalling USD 260 million during November, which is
significantly less than the USD 876 million recorded in the daily balance sheets and reported to
NBM.

According to interviews with BEM employees undertaken by Kroll during on-site visits, for each
deposit made by a Russian bank, it would have been necessary to produce a hard copy written
contract that needed to be posted to Russia in order to be signed by the Russian bank’s
representatives. Kroll was informed that this process ordinarily would take a minimum of 2-3 days
for the original documents to be returned from Russia. The fact that the large loans issued in
November 2014, guaranteed by the collateral agreements with Russian banks, were being issued
daily on multiple occasions suggests that it is unlikely that the deposits could have been agreed
(as per the process described above) before the loan decision was made.

Taken together, these factors provide evidence that the deposits that were used as collateral for
loans did not exist within the Three Moldovan banks, and were thus likely to have been
misrepresented to the NBM to present an illusion of liquidity.

4.2.3 Coordinated management of liquidity between the Three Moldovan


Banks

The loans repaid during the period 1-24 November 2014, mainly at UB and BS, which were
purportedly financed by the illusory Russian bank deposits at BEM, generated additional liquidity
at these two banks. During November 2014, this liquidity allowed BS and UB to deposit large
amounts of funds with BEM, which in turn allowed BEM to grant further loans to Shor Group
companies. This scheme involved the coordination of thousands of transactions across the Three
Moldovan Banks as well as related accounts in Latvia (detailed further in section 6.2). A summary
of the mechanism is shown in Figure 3:

9
The BEM Nostro account is an internal accounting record of the cash movements with other banks

Private and Confidential 29


Project Tenor II report

Figure 3: Interbank deposits increase liquidity in BEM and UB/BS

In interviews with Kroll, a representative of UB management stated that the instructions to place all
funds received from loan repayments on deposit at BEM were received directly from the Board of
Directors of UB (the “Board”). Senior management at the Three Moldovan Banks stated that the
negotiations for large interbank deposits were handled by the respective Board, and the managers
did not regularly discuss interbank deposits with their counterparties at the other banks. Examples
of how this process worked are provided in Section 4 of the Detailed Report.

As shown in Figure 3, the deposits originated in part from the loan repayments made by BEM to
UB and BS, which were linked to the increased liquidity at BEM provided by the Russian interbank
deposits. Additional details of the interbank deposits placed at BEM by BS and UB are included in
the Detailed Report.

4.2.4 Loan documentation and approval

Employee processing of loans

According to the employees interviewed at BEM, management of BEM instructed them orally on a
daily basis to execute the issuing of specific loans. With regards to the granting of customer loans
at BEM in November 2014, all BEM employees interviewed stated that they were following the
standard processes in place, 10 executing the duties required of their job and carrying out
instructions provided to them by BEM’s management and the BEM Board.

10
The Detailed Report sets out the detailed procedures for issuing of loans, as understood from on-site interviews.

Private and Confidential 30


Project Tenor II report

Employees interviewed stated that the issuing of loans required processing and input from four
departments at BEM: Credit, Treasury, Foreign Operations and Accounting. Each department 11
relied upon documents signed by the BEM Board for the execution of their duties. Although the
departments interacted with each other and exchanged information relating to the issuing of the
loans, staff did not challenge the information provided and relied upon the information provided by
other respective departments.

Documents reviewed by Kroll, supported the statements given by employees. Internal BEM
communications indicate that senior management was well aware of the massive increase in
lending activity taking place inside the bank. On repeated occasions, BEM employees from the
Credit or Treasury departments forwarded information to management of the bank to inform them
about the loans being issued. In a number of instances, these communications included
recommendations and warnings by BEM employees that the sudden increase in lending could
impact the institution’s financial stability, as well as informing management that documentation was
missing for a number of the loan applications. No responses from senior management were
identified within the electronic data set reviewed, which is consistent with the claims of staff during
interviews that they were provided with instructions verbally rather than in writing and that their
concerns were constantly ignored.

Board approval of loans

Employees at BEM stated in interviews with Kroll that in November 2014, all the decisions to
approve customer loans above the amount of MDL 1 million (USD 70 thousand) were made solely
by the BEM Board. They stated that this process was implemented following instructions from the
NBM that higher risk loans should also be approved by the Board. 12 Documents reviewed in the
recovered electronic data indicate that members of the BEM Board were informed of the risks
involved in the issuing of the loans to Shor Group companies during the November 2014 period.
There were several documents in hard copy that needed to be approved and signed by the BEM
Board in order for loans to be issued by the bank. 13

Notwithstanding the recommendations provided by BEM’s Credit Department, indicating a high risk
and likelihood of non-repayment by Shor Group companies, loans were repeatedly approved at
BEM Board meetings throughout November 2014.

Review of documentation revealed the following with regard to BEM board meetings:

11
All heads of department were interviewed with the exception of the accounting department head.
12
Kroll reviewed the BEM credit policy dated 30 December 2013 and effective for the 2014 year. Section 12 of this
Memorandum states that certain items need to be approved by the board of BEM in terms of regulation published by the
NBM on 23 October 2012, including all related exposures of MDL 1 million or above. This NBM regulation was
referenced as HCA-BNM no.238 in the BEM credit policy for the 2014 year. It is unclear from this document whether
BEM was applying the NBM regulation (HCA-BNM nr 238) prior to 2014.
13
Kroll understands that these documents were included in the batch of documents destroyed in the fire in transit
detailed in section 4.2.1. No scanned versions of these documents were identified during the electronic review.

Private and Confidential 31


Project Tenor II report

• There were two representatives from the Moldovan ministries of Finance and Economics
respectively appointed to the board of BEM. It appears that these government appointed
board members of BEM were frequently by-passed as they were not invited to board
meetings in November 2014. When contacted by Kroll to confirm their role, they stated that
they were not aware of any loans being granted by BEM in November 2014;

• Mr Shor was the Chairman of the Board of BEM from May 2013 to the end of 2014;

• During November 2014, a board log identified in the electronic data set suggests that
meetings took place almost daily (more frequently than would normally be the case) in order
to approve the high volume of loans;

• The purpose of these meetings appeared to be to approve all necessary operations to


continue to issue loans to the Shor Group companies; and,

• In addition to formal loan approvals, it was necessary to make decisions to secure interbank
deposits as collateral for loans (see 4.2.2), amend internal risk policies in order to increase
the lending ability of BEM, and accept requests by BS or UB to place overnight deposits.
These actions had the effect of facilitating the large increase in lending by BEM.

Further details about the composition of the BEM Board during November 2014 are detailed in the
Detailed Report, including examples of minutes of meetings recovered from the electronic data
review.

Escalation / Reporting

As noted above, employees interviewed stated that they believed they were performing their duties
as required, as the risks were highlighted in their submissions to the BEM Board. The electronic
documentation reviewed demonstrated that members of the Credit Department were escalating
risks to the organisation’s senior management.

It would appear that senior management and the Board repeatedly ignored such warnings and
granted the loans nonetheless. Detailed examples of the escalation of these risks by employees
are provided in the Detailed Report.

4.2.5 Coordinated loan applications at the Three Moldovan Banks

A review of electronic documentation obtained from the Three Moldovan Banks has identified that
throughout 2014, employees involved in the loan issuing process (mainly the Credit Departments)
at BEM and BS were communicating via email with representatives of the loan applicants, whose
email addresses contained the domain “@shorholding.com”. Despite this, BEM employees stated
that they did not suspect that the borrowing companies were legally owned by Mr. Shor, as it was
not evident that these representatives were also the owners of the borrowing companies.

Private and Confidential 32


Project Tenor II report

Further details of communication between employees at the Three Moldovan Banks and individuals
using a “@shorholding.com” e-mail address are provided in the Detailed Report.

4.3 Switch of the entire loan portfolio on 25-26 November


from BEM to BS

The Scoping Phase report identified that the entire loan portfolio issued by BEM during November
2014 was transferred to BS between 25 and 26 November 2014 (“the Switch”).

Following on from this preliminary analysis, Kroll traced the movement of funds involved in the
Switch and confirmed the involvement of at least 17 accounts held at a Latvian bank that acted as
a transit for the funds between Shor Group companies at BEM and BS. Loans totalling USD 956
million were issued by BS on the 25 and 26 November 2014 to five companies in the Shor Group.
These loan funds were then transferred to five accounts held at a Latvian bank.

From these accounts, the funds were moved through a network of other accounts held at the
Latvian bank, in an apparent attempt to disguise the fund flows. The laundering mechanisms for
the Switch transaction included issuing a “technical overdraft” facility by the Latvian bank to one
company, with a corresponding amount being simultaneously repaid to the Latvian bank from
another purportedly unrelated account. The funds were then used to repay loans at BEM totalling
USD 916.4 million. In addition, USD 39.2 million was transferred to BS from one of the Latvian
accounts. This was transferred to the account at BS, before being transferred to an account in the
name of Fortuna United LP, the company which purchased the entire loan portfolio. Details are
provided in Figure 4 below.

Private and Confidential 33


Project Tenor II report

Figure 4: Details of fund flows to switch the portfolio from BEM to BS on 25 - 26 November 2014

Private and Confidential 34


Project Tenor II report

4.3.1 Fake overdraft created in BS

The issuing of the loans in BS was only made possible by the manipulation of the accounting system
within BS and the creation of a fake overdraft facility on the BS correspondent bank account at the
Latvian bank. This was recorded only on the books of BS, as confirmed by both the employee at
BS who was instructed to do this and the Latvian bank itself (further details are provided in the
Detailed Report). This was only possible because the funds all passed through the correspondent
account within one business day and did not therefore require an overnight balance on the
correspondent bank account, which would have been subject to daily clearing by the Latvian bank.

4.3.2 Reallocation of interbank deposits


According to the balance sheets submitted to NBM by the Three Moldovan Banks, between 24
November 2014 and 27 November 2014 the interbank deposits and cash of BEM and UB were
consolidated into BEM term deposits held at BS. In addition, as BS loans were used (after being
subjected to a laundering mechanism) to repay loans at BEM, BEM increased its funds placed on
deposit at BS. This resulted in a large increase in the BEM term deposits at BS line in the balance
sheet of BS between 24 November and 27 November, as detailed in the Detailed Report.

Private and Confidential 35


Project Tenor II report

Shor Group connections with the ownership of


the Three Moldovan Banks
Section 4 demonstrated that the increase in the loan portfolios of the Three Moldovan Banks was
made possible through the central control and co-ordination of the Three Moldovan Banks. It
appears that this level of control was possible as parties connected to the Shor Group also became
owners of the Three Moldovan Banks through an opaque offshore structure.

The Three Moldovan Banks underwent a change in ownership in 2012 (UB) and 2013 (BS and
BEM). During the Scoping Phase of work, a preliminary analysis was undertaken of the
shareholding structure of BEM, UB and BS, in order to identify whether there was any correlation
between the ownership structures, the control of the Three Moldovan Banks and the likely
beneficiaries of the transactions. The preliminary conclusion at that stage was that the new
shareholders of the Three Moldovan Banks were acting as nominees, apparently put in place to
deliberately disguise the true beneficial ownership of shares and that a number of shareholders
financed their acquisition with loans from UK Limited Partnerships, whose ultimate beneficial
ownership was unclear and who had accounts at Latvian banks, through which the funds were
transferred.

With regards to BEM, the change in ownership involved a cession agreement that transferred a
number of legacy non-performing loans to a third party. This was signed between BEM and a UK
partnership called Roseau Alliance LLP (the “Roseau Transaction”) whereby approximately USD
80 million of non-performing loans were purchased by Roseau Alliance LLP (“Roseau”) from BEM
2013.

Kroll has significantly progressed the understanding of the fund flows that allowed the coordinated
acquisition of the Three Moldovan banks, and has identified that the source of the funds to purchase
shares came from both Shor Group company loans issued by the Three Moldovan Banks, 14 and
from accounts held at Russian banks. Furthermore, the accounts held at one Russian bank that
funded the acquisition of the shares in UB or BS, or funded the cession agreement for BEM were
held in the name of Shor Group companies, or individuals associated with such companies or Shor
himself. Additionally, analysis of transactional data has revealed that the funds which originated
from the Russian banks appeared to have been ultimately financed by loans issued by the Three
Moldovan Banks. Details of this are set out in the Detailed Report. Further details are provided
below in relation to the particular banks.

14
Funds originating from Shor Group company accounts came in part from loans issued by the Three Moldovan Banks.

Private and Confidential 36


Project Tenor II report

5.1 Unibank (UB)


The Scoping Phase report stated that from 2007 until August 2012, UB was under the ownership
of Austria-based Vienna Capital Partners Unternehmungsberatungs AG. On 17 August 2012, all of
the bank’s shares were sold and transferred to 21 new shareholders, each with a stake between
4.5% and 4.99%. 15 A shareholder who held a stake of at least 5% was classified as a significant
shareholder, with their acquisition subject to formal approval by the NBM. The process was
therefore circumnavigated by this scheme.

The Scoping Phase identified that this group of 21 shareholders were likely to be holding shares
as nominees on behalf of an undisclosed beneficiary and that the fund flows to purchase the shares
came from a series of Latvian accounts, many in the name of UK Limited Partnerships. At the time
sufficient information was not available to trace these funds. During the subsequent phases of work,
Kroll has obtained statements for accounts of numerous linked companies held at two Latvian
banks, and has traced the source of shareholders’ funds to the extent possible.

Kroll has extended the analysis from the Scoping Phase and has identified USD 3.1 million (in
funds from accounts held at two Russian banks in the name of companies or individuals linked to
the Shor Group) and USD 11.3 million (from loans from the Three Moldovan Banks) as the original
source of funds for these share purchases. In addition, Kroll has identified USD 2.1 million used to
finance the share acquisition that originated from Shor Group companies with accounts at other
Moldovan banks. As with the transfer and dissipation of funds detailed elsewhere in this report,
numerous interim accounts were used to merge and layer funds, to mask the true origin of the
funds.

The details of the share acquisitions that were financed through this channel and the related fund
flows are set out in the Detailed Report. Figure 5 shows an example of one of the fund flows to
acquire UB shares, funded by a combination of funds originating from a Shor Group company
account at a Russian bank, and a loan to a Shor Group company.

15
According to NBM data, which showed no change in shareholding data, irrespective of additional shares issued by the
bank.

Private and Confidential 37


Project Tenor II report

Figure 5: Illustration of the funding of the purchase of UB shares

(All values in USD thousands)

This figure demonstrates a typical example of a combination of funds that originated from Shor
Group companies at other Moldovan banks and Shor Group companies or linked parties at Russian
banks before they were mixed and transferred through multiple accounts at Latvian banks and
ultimately transferred to the account of a shareholder at UB around the date they became a
shareholder of UB. Further details of all such shareholders can be found in the Detailed Report.

5.2 Banca Sociala (BS)


The Scoping Phase report highlighted that the shareholding of BS changed in May 2013. Before
May 2013, BS was principally controlled by three groups, with a large number of minority
shareholders making up the remainder of the ownership. As a result of the sale of over 50% of the
bank’s share capital between May 2013 and June 2013, the shareholding of BS changed
significantly, with no single shareholder holding above 5%. These new shareholders comprised

Private and Confidential 38


Project Tenor II report

Moldovan companies, and Moldovan, Russian and Ukrainian individuals. The work conducted in
the Scoping Phase identified a number links between these new shareholders. At that stage it was
identified that the funds used to purchase the shares were all received from a group of bank
accounts at three Latvian banks.

The work progressed since the Scoping Phase has established the source of USD 6.1 million of
funds involved in the acquisition of BS as being from accounts held in the names of a number of
individuals who were linked to Shor Group. Kroll’s analysis has traced the source of a further USD
11.6 million. The source of funds that were used to acquire shares in BS was as follows:

• Funds originating from Shor Group loans issued by the Three Moldovan Banks – USD 11.6
million;

• Funds originating from Shor Group or other accounts linked to Shor Group companies at
Russian Banks – USD 1.1 million; and,

• Funds traced to an account at a Latvian bank that was not part of the primary laundering
mechanism detailed below – USD 5 million.

The details of the shareholders who were financed through this channel and the related fund flows
are included in the Detailed Report.

5.3 Banca de Economii


Subsequent to the change in share ownership structure at UB and BS, BEM also underwent a share
restructure in 2013.

The Scoping Phase report identified that prior to the share restructure, on 18 March 2013, BEM’s
management signed a cession agreement, 16 which resulted in over MDL 1 billion (USD 80 million)
of non-performing loans being removed from BEM’s balance sheet in exchange for cash payments
of the same value, which significantly increased liquidity at BEM. This resulted in the transfer of a
series of non-performing loans to Roseau.

Kroll traced the funds used to acquire these loans and established that a proportion of funds came
from bank accounts at a Russian bank, as well as from loans to Shor Group Company accounts
issued by the Three Moldovan Banks.

16
The agreement was signed by senior management of BEM

Private and Confidential 39


Project Tenor II report

At the time of the Scoping Phase report, the bank statements provided to Kroll were not sufficient
to trace more than USD 6.3 million of the USD 81.6 million in proceeds 17 used to purchase the non-
performing loans. This analysis has been continued and extended in the current phase to allow a
greater understanding of the Roseau Transaction. Kroll’s detailed analysis has identified
USD 58.5 million used to purchase the non-performing loans book sourced from bank accounts
held at Russian Bank 1 by 13 named individuals, many linked to the Shor Group and a further
USD 21.1 million was sourced from Moldovan bank accounts held by 16 named entities also linked
to the Shor Group. These funds were in part derived from loans to Shor Group companies.

The Detailed Report details and illustrates the fund flow for each of the individuals and entities that
provided funds to purchase these non-performing loans.

17
The traces detailed in the Detailed Report total USD 79.6 million. The difference from the USD 81.6 million is due to a
different exchange rate used in the Scoping Report.

Private and Confidential 40


Project Tenor II report

Laundering of loan funds


Kroll has progressed the tracing of loan proceeds issued to Shor Group companies by the Three
Moldovan Banks. The Scoping Phase report identified that the majority of the funds were
transferred to accounts at Latvian banks and that a proportion of the funds appeared to be used to
repay existing loan balances, having been passed through Latvian accounts and having been
transferred back to Moldova. NBM has provided significant disclosure, which has been
supplemented with disclosure to the NBM by authorities in Latvia and Estonia. Kroll has used this
information to progress the analysis of fund flows, to develop a priority recovery strategy for funds
that have been dissipated to other locations. A summary of the findings is set out in the sections
below. Specifically:

• An analysis was carried out of the initial destination of loan funds issued to Shor Group
companies by the Three Moldovan Banks between January 2012 and the end of November
2014. This analysis identified that USD 2,617.9 million (or 89% of all) of loan funds issued
to Shor Group companies was transferred direct from the current accounts that received
the loan, to accounts at two Latvian banks mainly in the name of UK limited partnerships
or other companies with opaque ownership structures in offshore secrecy locations. The
remainder of the funds were transferred to accounts in other jurisdictions, including
accounts at Russian banks, other Moldovan banks and accounts at other Latvian banks.
Further details are provided in section 6.1;

• Further analysis of the amounts transferred to accounts at the two Latvian banks identified
that the loan funds were transferred through a group of interconnected accounts that
displayed signs of being centrally controlled (the “Core Laundering Mechanism”). A
significant portion of loan funds (USD 2,015.3 million) was returned to accounts at the Three
Moldovan Banks to repay existing loans. A number of additional laundering mechanisms
were also identified. Further details of these laundering mechanisms are included in section
6.2; and,

• Further forensic funds tracing of the dissipated loan funds out of the Core Laundering
Mechanism revealed that a significant portion of these funds were transferred to bank
accounts in other jurisdictions, including Cyprus (USD 112.4 million), China (USD 83.3
million), Russia (USD 80.6 million), Switzerland (USD 42.8 million), USA (USD 25.5 million)
and Hong Kong (USD 22.6 million). Additional details of the destination of funds are
included in section 7.

Private and Confidential 41


Project Tenor II report

6.1 Initial destinations of loan funds


Kroll analysed all of the loans issued to Shor Group companies between January 2012 and the end
of November 2014, to clarify the initial destination of loan funds issued, 18 as shown in Figure 6:

Figure 6: Initial dissipation of loan funds

18
Loans included in this analysis were selected in accordance with the methodology documented in the Detailed Report.

Private and Confidential 42


Project Tenor II report

Table 6 summarises the initial destinations of suspected fraudulent loan funds issued by the Three
Moldovan Banks between January 2012 and November 2014. 19 The majority of these loans were
laundered through the Core Laundering Mechanism and returned to Moldova to repay existing
loans.

Table 6 : Initial destination of loan funds 2012 – November 2014

Initial destination Amount (USD %


million)
Paid to Latvian Bank 1 477.5 16

Paid to Latvian Bank 2 2,140.4 73

Subtotal 20 2,617.9 89

Other jurisdiction 106.2 4


Remained in Moldova 218.7 7
TOTAL 2,942.8 100

6.1.1 Funds to the Core Laundering Mechanism


Two Latvian banks received the majority of the proceeds of the suspected fraudulent loans. Kroll
identified that prior to May 2014, transactions totalling USD 477.5 million were transferred to
accounts held at Latvian Bank 1. Many of the accounts that received the funds at Latvian Bank 1
were held in the name of UK limited partnerships, or limited companies registered in offshore
jurisdictions, such as BVI, Seychelles and Belize. Significant transfers during this period were also
made to Moldovan accounts and accounts at Russian banks.

The majority of the loans issued after May 2014 (which made up the largest proportion of the volume
of loans issued) totalling USD 2.1 billion were transferred to accounts at Latvian Bank 2,
predominantly to accounts held in the names of UK limited partnerships.

The Core Laundering Mechanism was made up of a series of accounts which transacted
predominantly with each other and accounts at the Three Moldovan Banks. Further details of this
are provided in section 6.2.1 and in the Detailed Report.

19
This includes the “Switch” transaction, where USD 1 billion of loans was issued by BS to repay loans consolidated at
BEM between 25 and 26 November 2014.
20
There is a difference of USD 35.6 million between the total paid to Latvian Bank 1 and Latvian Bank 2 in Table 6 and
Table 7. This is because USD 35.6 million was paid to accounts at these two banks that were not included as Core
Laundering Accounts. These accounts were classified as ancillary accounts. Additional detail is included in the Detailed
Report. The amount paid to Core Laundering Accounts is USD 2,582.3 million.

Private and Confidential 43


Project Tenor II report

6.1.2 Destination of funds that did not pass through the core laundering
mechanism

Amounts paid to non-Core Laundering Accounts at Latvian Bank 1


and Latvian Bank 2
In addition to the funds transferred to the accounts classified as part of the Core Laundering
Mechanism (the “Core Laundering Accounts”), Kroll also identified a further USD 35.6 million
transferred to other accounts at these banks . Kroll has traced these amounts onwards in
accordance with the funds tracing methodology detailed in the Detailed Report.

Amounts transferred to other jurisdictions


USD 106.2 million originating from new loans issued by the Three Moldovan Banks between
January 2012 and November 2014, was transferred to accounts in jurisdictions other than Moldova
and the two Latvian banks involved in the Core Laundering Mechanism.

Amounts transferred to Russia (Total USD 97.8 million)

A total of USD 97.8 million was transferred to accounts held at Russian banks. Further details of
these fund flows can be found in section 6.2.3 and in the Detailed Report.

Amounts paid to other jurisdictions (Total USD 8.4 million)

A total of USD 8.4 million that originated from loan funds from the Three Moldovan Banks to related
companies was transferred direct to other jurisdictions without entering the Core Laundering
mechanism. These amounts were paid mainly to accounts at other Latvian banks and are detailed
in the Detailed Report.

Amounts remaining in Moldova

A proportion of the loan funds that were issued to Shor Group companies by the Three Moldovan
Banks was not transferred overseas. Analysis indicates that funds were transferred to accounts at
various Moldovan banks (including the Three Moldovan Banks), to pay down existing loan exposure
or to fund (what appears from the transaction descriptions) trading activity.

Loan funds used directly to repay earlier loans at Three Moldovan Banks

Analysis has identified a pattern of transactions that occurred during 2012 and 2013, particularly
relating to loans that were issued by UB, whereby loan funds were mixed and used to repay earlier
larger loans. A number of transfers were made to accounts at BS and BEM during this period.

Loan proceeds transferred to other Moldovan banks

Private and Confidential 44


Project Tenor II report

A proportion of the loan proceeds that remained in Moldova were transferred to accounts at seven
Moldovan banks. Further details are included in the Detailed Report. It is currently not clear whether
the transfers made to these Moldovan banks were part of further laundering activity, or whether
they represent an end destination of funds. It appears however from analysis to date that accounts
in two banks which received the majority of the funds in Moldova were also involved in the
coordinated issuing of loans to Shor Group companies that were serviced and/or repaid by loans
issued by the Three Moldovan Banks.

Kroll does not yet have access to the full data set at other Moldovan banks. The fund flows analysed
to date, however, indicate that the actual dissipation of funds may have occurred earlier than the
Review Period from loans issued by other Moldovan banks, which were subsequently repaid by
loans issued by the Three Moldovan Banks. Further analysis will be required to clarify whether
these funds were dissipated earlier for the benefit of particular individuals or whether they
represented genuine loans that subsequently defaulted and were then funded by new loans issued
by the Three Moldovan Banks to cover up the defaults.

6.2 Laundering mechanisms


As detailed in section 6.1, USD 2,582.3 million was transferred to a group of accounts at two Latvian
banks, which together with the accounts used to transit the funds made up the Core Laundering
Mechanism. Kroll has identified that between January 2012 and April 2014, Latvian Bank 1
laundered the funds that were issued by the Three Moldovan Banks. From May 2014 to the end of
November 2014, the Core Laundering Mechanism was changed to Latvian Bank 2. In addition to
the Core Laundering Mechanism, which appears to have been established solely for the purpose
of laundering the Moldovan loan funds, Kroll has also identified additional peripheral laundering
mechanisms in Moldova, Estonia, other Latvian banks, and possible further laundering
mechanisms in other jurisdictions.

6.2.1 Core Laundering Mechanism


The Core Laundering Mechanism as detailed above involved the movement of huge sums (by
amount and by volume of transactions) through the Core Laundering Accounts have all of the key
characteristics of money laundering, including the placement, layering and subsequent integration
of funds.

Core Laundering Accounts


Kroll analysed hundreds of bank statements of companies that received and disbursed the loan
funds at both Latvian Bank 1 and Latvian Bank 2. Preliminary analysis of these accounts revealed
that similar patterns were present in both the Latvian Bank 1 and Latvian Bank 2 laundering
mechanisms. These common characteristics exhibited by the funds transfers included multiple

Private and Confidential 45


Project Tenor II report

internal transfers to linked accounts, splitting and merging of funds and multiple foreign exchange
transactions. This mechanism is demonstrated in Figure 7.

Further details about these characteristics are included in the Detailed Report.

Private and Confidential 46


Project Tenor II report

Figure 7: Overview of core laundering mechanisms in Latvia

Latvian Bank 1

Latvian Bank 2

Private and Confidential 47


Project Tenor Phase II Report

The majority of the Core Laundering Accounts at Latvian Bank 1 and Latvian Bank 2 that received
and laundered the funds were held by UK registered Limited Partnerships. There were also
accounts registered to companies in other offshore secrecy locations including Gibraltar, Belize,
BVI and the Marshall Islands. These accounts appear to have been opened for the sole purpose of
laundering the proceeds of the fraudulent loan transactions from the Three Moldovan Banks.

Although the value of the loans transferred and laundered through Latvian Bank 2 was significantly
greater than that which went through Latvian Bank 1, the Latvian Bank 2 mechanism involved fewer
companies and was over a shorter period of time.

Many of the bank accounts held by UK Limited Partnerships at Latvian Bank 2 are linked to the
Shor Group. In particular, account opening documentation established that Mr Shor was present at
the bank in Latvia and made representations about the stated business purpose of the accounts as
a group. This information, along with the way the transactions between the accounts was co-
ordinated, strongly suggest that these accounts were acting in concert and were being centrally
controlled. Additional information about the links between these companies and the Shor Group is
included in the Detailed Report. Kroll has not yet been given access to similar documentation at
Latvian Bank 1, which made up the Core Laundering Mechanism. However, as the process and
mechanisms appear to have many similar characteristics and received loan funds from similar
companies, it appears that they also were part of the same centrally co-ordinated laundering
process.

Private and Confidential 48


Project Tenor Phase II Report

Summary of all funds into the Core Laundering Accounts

Table 7 : Summary of all funds received into the Core Laundering Accounts
Source of funds Latvian Bank 1 Latvian Bank 2 – Total – (USD
(USD millions) (USD millions) millions)

Receipts to the Core Laundering 447.5 2,134.8 2,582.3


Accounts from new loans issued

Receipts from Shor Group companies 88.6 8.9 97.5


that could not be directly linked to loans

Less amounts on Moldovan statements - (101.0) (101.0)


that did not appear on the Latvian Bank
2 accounts (See the Detailed Report
and Section 6.2.4)

TOTAL RECEIPTS to Latvian Bank 1 / 536.1 2,042.7 2,578.8


Latvian Bank 2 from BEM / UB / BS
from Shor companies

TOTAL OTHER RECEIPTS (not from 174.9 126.5 301.4


BEM / BS / UB) 21

TOTAL RECEIPTS 711 2,169.2 2,880.2

Other funds received on the Core Laundering Accounts


As noted above, the primary purpose of the Core Laundering Accounts appeared to be to disguise
and to launder the loan funds that originated from the Three Moldovan Banks. In addition, as
highlighted in the table above, a total of USD 301.4 million of other funds was identified as entering
the Core Laundering Accounts (from destinations other than the Three Moldovan Banks or the Core
Laundering Accounts themselves). These funds were frequently mixed with funds originating from
the fraud at the Three Moldovan Banks, and then dissipated. As such, these transfers were treated
as being linked to the suspected fraud in whole or part. In certain cases, funds were not mixed and
could be excluded from the analysis of onward dissipation. As detailed in Table 8, the amount of
other receipts that could be separated and excluded from the fraud funds amounted to USD 157.8
million. These amounts were not included in the analysis of dissipated ends in section 7 as neither
their source nor their destination could (at this stage) be associated with the fraud.

21
Includes amounts transferred from other Moldovan banks, non- Core Laundering Account at Latvian Bank 1 / Latvian
Bank 2, other Latvian banks and accounts in other jurisdictions. Excludes receipts from Russian accounts, which is
analysed separately in the Detailed Report.

Private and Confidential 49


Project Tenor Phase II Report

Table 8: Other Receipts traced out


Source of funds Latvian Bank 1 Latvian Bank 2 – Total – (USD
(USD millions) (USD millions) millions)

TOTAL RECEIPTS to Latvian Bank 1 536.1 2,042.7 2,578.8


/ Latvian Bank 2 from BEM / UB / BS
from Shor Group companies

TOTAL OTHER RECEIPTS (not from 174.9 126.5 301.4


BEM / BS / UB)22

Less funds traced out (75.9) (81.9) (157.8)

TOTAL OTHER FUNDS MIXED 99 44.6 143.6

Total funds to trace out of the Core 635.1 2,087.3 2,722.4


Laundering Accounts

Further details of other receipts and the treatment of these transfers is included in the Detailed
Report.

6.2.2 Fund flows involving accounts at Russian banks


Kroll noted further involvement of accounts at Russian Banks in the transfer of funds connected to
the fraud at the Three Moldovan Banks.

Two distinct schemes were identified involving transactions between accounts at the Three
Moldovan Banks, the Core Laundering Accounts and accounts at three different Russian banks. In
both of these schemes, funds were transferred from Russian accounts to accounts at the Three
Moldovan Banks or the Core Laundering Accounts. These funds were then transferred to the
Russian banks at a later date.

The circular movement of funds from accounts at the Russian banks to the Core Laundering
Accounts that were subsequently repaid confirms that these accounts appear to have been used
to add an additional layer to disguise the flows of funds and to allow lending to continue through
providing liquidity, as well as enabling BEM to misrepresent its current liquidity to NBM through the
misstatement of the current interbank balances. This was particularly prevalent in relation to
movements with the same two Russian banks that were purported to have issued the questionable
Russian Bank deposits, as the majority of the fund flows were to and from accounts held in the

22
Includes amounts paid from other Moldovan banks, non- Core Laundering Accounts at Latvian Bank 1 / Latvian Bank
2, other Latvian banks and accounts in other jurisdictions. Excludes receipts from Russian accounts, which is analysed
separately in the Detailed Report.

Private and Confidential 50


Project Tenor Phase II Report

name of Shor Group companies, or individuals connected with the Shor Group. Details of the
specific transactions and links to Shor Group companies can be found in the Detailed Report.

In addition to the funds that appear to have been repaid, Kroll also identified further outflows of
funds to Russian banks, which are detailed in the Detailed Report and summarised in section 6.2.3.

The two schemes are summarised below, and are detailed further in the Detailed Report:

Funds from two Russian banks, repaid using loans issued to Shor Group companies

Analysis of the Core Laundering Accounts identified a series of transfers totalling USD 199 million
into the Core Laundering Mechanism from accounts at two Russian banks. The transfers were
from a number of accounts that appeared to be linked to the fraud, as they were in the names of
Shor Group companies, individuals linked to the Shor Group (for example the directors of Shor
Group companies) or other Moldovan individuals. The transfers into the Core Laundering
Mechanism were made during 2012 and 2013 when the funds in the Core Laundering Mechanism
were being laundered by Latvian Bank 1. In 2014, the funds appear to have been largely repaid to
the two Russian banks from Latvian Bank 2, using funds that could be traced to loans issued by
the Three Moldovan Banks. Transfers were identified totalling USD 193.7 million to accounts at the
two Russian banks that also appeared to be associated with the Shor Group or other individuals
linked to Shor. Further details can be found in the Detailed Report.

Scheme involving Russian Bank 3


Toward the end of 2014, funds from a third Russian bank (“Russian Bank 3”) appear to have been
used as part of a scheme to inflate the reported liquidity of BEM as disclosed to NBM. .

Funds were received into accounts at BEM from accounts at Russian Bank 3 as part of a
subordinated loan agreement and for the purchase of shares by one of the major shareholders of
BEM, Sisteme Informationale Integrate (“SII”). According to BEM’s records, these amounts
(transferred to BEM) were placed on overnight deposit at Russian Bank 3 until November 2014.
However, the account records for SII show that USD 50 million of the funds were in fact transferred
out to the Core Laundering Mechanism and were not returned to the SII account or repaid to the
account at Russian Bank 3.

SWIFT records reviewed from September 2014 show that USD 100 million was transferred to the
accounts at Russian Bank 3 by SII and from BEM’s other shareholder accounts. However, BEM
continued to show the cash deposit at Russian Bank 3 on its nostro account for a further two
months. This amount was then cleared, through the apparent transfer of USD 100 million from two

Private and Confidential 51


Project Tenor Phase II Report

Shor Group companies to companies in the Core Laundering Mechanism, which were not recorded
on the statements of the Core Laundering Accounts.

6.2.3 Overview of amounts dissipated to Russia


Analysis of the schemes used to disguise the flow of funds through the Russian banks indicates
that not all funds transferred to Russia are in fact the end dissipation of fraud funds, as many of
these funds appear to be repayments of amounts previously received from Russian accounts.
However, Kroll’s analysis has identified that there was a net outflow of USD 47 million transferred
or received direct from the Core Laundering Accounts and the Three Moldovan Banks.

Analysis of the scheme involving Russian Bank 3 identified a net inflow of USD 3.7 million from
accounts in Russia, apparently used for the purchase of shares in BEM.

The further tracing exercise, detailed in section 7 revealed an additional USD 37.5 million to other
Russian banks identified through forensic funds tracing that was transferred to other accounts at
banks in Russia.

As shown in Table 9 this amounted to a net outflow of USD 80.6 million to accounts held by banks
in Russia that originated in part from the loans from the Three Moldovan Banks.

Table 9: Net effect of transactions with Russian banks:


Source of funds Russian Bank(s) Net amount transferred
to /(received from)
Russian accounts
(USD million)
Transferred direct from loans from the Three Russian Bank 1, 46.8
Moldovan Banks Russian Bank 2
BEM Russian Bank 3 (3.7)
Further tracing Several Russian 37.5
banks
Total dissipated 80.6

6.2.4 Other laundering mechanisms identified

In addition to the Core Laundering Mechanism detailed above, Kroll also identified a number of
other secondary laundering mechanisms that are summarised below. Further detail of these is
provided in the Detailed Report, including detailed examples of fund flows.

Private and Confidential 52


Project Tenor Phase II Report

Laundering of funds through accounts at other Moldovan banks

In addition to the loan funds that were transferred to the two Latvian banks from the Three Moldovan
banks, a proportion of funds were subjected to a layer of laundering using accounts at other
Moldovan banks, prior to being transferred to the Core Laundering Mechanism, or were used to
repay existing loans. These transfers do not appear to have an obvious business rationale and
appeared to have been effected with the sole purpose to disguise the source of the funds before
entering the Core Laundering Accounts. Kroll has not identified any evidence at this stage of
complicity or involvement of management of other Moldovan Banks in the fraud.

Laundering of funds through Estonian banks and other Latvian


banks

Kroll identified a number of accounts at Estonian banks and other Latvian banks 23 that received
funds once they had passed through the Core Laundering Mechanism detailed above (see section
6.2.1). Analysis of the fund flows through these Estonian and other Latvian accounts has shown
that a number of these accounts appear to be linked and part of a co-ordinated approach to further
disguise the flow of funds.

Other laundering systems

China / Hong Kong / Cyprus

Further tracing has identified that a significant proportion of the funds were paid to banks accounts
in the People’s Republic of China (“China”), Hong Kong and Cyprus. The transfers to Hong Kong
and China appear to be to accounts in the name of local companies, while a number of the Cyprus
recipients appear to have similar characteristics to the companies in the Core Laundering
Mechanism, i.e. being either UK limited partnerships or companies registered in offshore
jurisdictions. These transactions have descriptions that indicate that they are for trade purchases
(such as construction materials, textiles, furs or shoes). As we do not currently have access to the
statements for the accounts held at the Chinese, Hong Kong or Cypriot banks, it is not possible to
determine whether these transactions are for genuine purchases or whether they represent an
entry point to another layer of money laundering. Further disclosure is necessary to determine the
ultimate destination of amounts paid to these accounts.

23
Latvian banks other than those involved the Core Laundering Mechanism

Private and Confidential 53


Project Tenor Phase II Report

End destination of fraudulent loan funds


Based on documentation reviewed to date, the proceeds of the loans that were subject to
laundering through the Core Laundering Mechanism and other laundering mechanisms detailed in
the sections above were traced to establish the final destination of funds as far as possible, and to
attempt to identify assets or beneficiaries to enable legal steps to be initiated. The detailed
methodology that was used to trace these funds is in the Detailed Report.

End destinations identified in the analysis of the initial dissipation of loan funds summarised in
section 6.1, and the funds flowing out of the Core Laundering Mechanism summarised in section
6.2.1 were aggregated to produce a consolidated table of the end jurisdiction of the funds
originating from the suspected fraudulent loans issued by the Three Moldovan Banks (summarised
in Table 11). Details of the account holders and banks to which funds were transferred are in the
Detailed Report.

To date, Kroll has identified USD 1.1 billion (including loan proceeds and funds mixed with the loan
proceeds) transferred to bank accounts in a number of jurisdictions, detailed in section 7.1.1. A
proportion of this amount is made up of funds that have been mixed with loan funds during the
tracing process, but for each transaction identified, it is possible to conclude that the final transfer
identified to date is made up at least in part from funds originating from the suspected fraudulent
loans to the Shor Group companies from the Three Moldovan Banks. This is summarised below:

Table 10: overview of total end destinations of funds

Description Amount Link to report


(USD section
billion)

Initial amount of loss 0.6 4.1.4

Add: Mixed funds from other laundering mechanisms 0.5 6.2.4

TOTAL 1.1 7.1.1

Private and Confidential 54


Project Tenor Phase II Report

7.1.1 Overview of end jurisdictions

The amounts traced to ends, either dissipated from loan funds at the Three Moldovan Banks, paid
direct from the Core Laundering Accounts, or through the further laundering mechanisms 24 are
listed in Table 11 below. As disclosure of the bank accounts for the ends currently identified has
not yet been obtained, the final destination of the funds which flowed to these accounts is currently
unknown. It is likely that the funds were transferred through these accounts to other destinations
or to purchase assets. It will be necessary to receive the cooperation of the relevant supervisory
authorities or obtain disclosure orders to further trace the dissipation of the stolen funds through
these and other jurisdictions, and to identify beneficiaries.

Table 11: Amounts traced to ends by jurisdiction


End jurisdiction of bank account Total value ends
Latvia (excluding Core Laundering Accounts) 302.1
Moldova (other Moldovan Banks) 169.3
Cyprus 112.4
China 83.3
25
Moldova (BEM / BS / UB) 82.9
26
Russia 80.6
27
Austria 56.1
Estonia 51.4
Switzerland 42.8
USA 25.5
Hong Kong 22.6
Italy 16.8
Romania 15.8
Germany 11.1
Turkey 10.4
Other jurisdictions (less than USD 10 million in aggregate) 50.5
Other jurisdictions (less than USD 1 million in aggregate) 10.9
28
TOTAL 1,144.6

Accounts in the following jurisdictions received funds of between USD 1 – USD 10 million in
aggregate: Netherlands, United Arab Emirates, Singapore, Liechtenstein, India, UK, Poland,

24
This table includes amounts dissipated directly from loan funds, amounts dissipated from Core Laundering Accounts
and amounts dissipated following further laundering.
25
Amounts dissipated to BEM / BS / UB were from traces after further laundering after being dissipated from the Core
Laundering Accounts.
26
See Detailed Report for details of other Russian transfers that are not included in the end dissipation.
27
This amount comprises the transfers used for the purchase of BEM’s non-performing loans under the cession
agreement with Roseau, detailed in section 5.3
28
Differences as a result of rounding

Private and Confidential 55


Project Tenor Phase II Report

France, Luxembourg, Czech Republic, Ukraine, Vietnam, Brazil, Ireland, Monaco, Andorra, South
Korea and Taiwan. Accounts in various other jurisdictions also received funds in aggregate of less
than USD 1 million.

7.1.2 Ends in accounts in Latvia and Estonia

A number of traces currently ended in Latvia, Estonia and Moldova are in progress, with account
statements pending. For the sake of completeness, these amounts have been included in the table
above, but these are likely to change if further analysis is conducted. Additional details are provided
in the Detailed Report.

Private and Confidential 56


Project Tenor Phase II Report

Identification of beneficiaries / perpetrators


A proportion of the fund flows which have been traced to date have revealed a number of individuals
who appear to have either been part of the administration of the suspected fraudulent activity, or to
have obtained some benefit from the outflow of the funds. This benefit was identified partly through
the flow of funds to companies with known affiliations to individuals. Receipts into personal
accounts or into companies known to be linked to these individuals totalled approximately USD 50
million.

As stated in the introduction to this report, in order to protect due process with regards to ongoing
or future civil or criminal procedures, it is important that the apparent beneficiaries are kept
confidential except in cases which can contribute to any ongoing procedures. Details will be
provided to the relevant authorities under separate cover.

Private and Confidential 57


kroll.com